AI/ML Engineer wanted for building AI for triage first responses

1. Strong experience in AI/ML & NLP (chatbots, LLMs, or sentiment analysis).
2. Proficiency in Python + frameworks like PyTorch/TensorFlow.
3. Ability to build & deploy prototypes into production (cloud or local).
4. Awareness of data privacy & security (GDPR/FADP).
5. Startup mindset: can work independently, fast, and resourcefully.

• Machine Learning
• Natural Language Processing (NLP)
• Large Language Models (LLMs)
• Python
• PyTorch / TensorFlow
• Hugging Face
• Speech-to-Text / Voice AI
• Sentiment analysis
• Cloud Deployment (AWS / GCP / Azure)
• Data Privacy (GDPR / FADP)
• Startup Mindset

You have strengths in...

• words and communication
• numbers, data and logic
• pushing and getting things done

You already have...

• exited a startup
• worked as fulltime employee

Build the first prototype (NLP model for triage, risk detection).

Set up voice/chat interface.

Work with open-source models + fine-tuning.

Equity / Co-Founder role: Be part-owner of SafeLine AI and shape the company from day one.

Meaningful impact: Build technology that ensures no psychiatric crisis call in Switzerland goes unanswered.

Creative freedom: Lead AI development, choose the stack, and influence the product roadmap.

Visibility & credibility: Work with early hospital/NGO partners, add “Founding AI Engineer” to your profile.

Future upside: As funding arrives (grants, investors), you’ll transition to a paid leadership role.

No psychiatric crisis call unanswered: AI triage that saves lives, supports clinicians, and builds trust in Switzerland’s mental health system

SafeLine AI offers Switzerland’s first AI-powered triage assistant for psychiatric crises — ensuring every call is answered, assessed, and safely escalated.

24/7 AI Crisis Triage
Always-on voice and chat assistant that ensures no psychiatric crisis call goes unanswered.

Risk Assessment & Escalation
Structured questions + AI sentiment analysis to detect urgency and safely escalate to 144, a clinician, or hotline staff.

Support for Clinics & Hotlines
Reduce missed calls, staff overload, and costly emergency hospitalizations by bridging the gap until a human responder is available.

Challenges

1. Regulation & Trust: AI in psychiatric care faces strict laws and clinician skepticism; adoption requires medical advisors and careful positioning.

2. Safety & Liability: False negatives in crisis detection carry high reputational and legal risk.

3. Funding & Focus: Long hospital sales cycles mean early reliance on grants, pilots, and a clear niche.

Target group

Primary Target Audience (the decision-makers who will pay)
1. Swiss Psychiatric Hospitals & Clinics:
a. Example: PUK Zürich, Universitätsspital Zürich.
b. Pain: too many missed crisis calls, staff overload, expensive hospitalizations.
c. Value: guaranteed triage, cost savings, reduced clinician burnout.

2. Hotlines & NGOs
a.Example: Dargebotene Hand (143), cantonal crisis hotlines.
b. Pain: can’t answer every call, reputation risk when people don’t get help.
c. Value: AI ensures no call unanswered, escalates only when needed.

3. Health Insurers (Krankenkassen)
a. Example: Helsana, CSS.
b. Pain: psychiatric hospitalizations are very costly.
c. Value: prevention saves money → AI triage reduces unnecessary admissions.

Secondary Target Audience (influencers & partners)
1. Cantonal Health Departments → public mandate to ensure crisis services.
2. Clinicians & Psychiatrists → key validators; need to trust your solution.
3. Impact Investors & Healthtech VCs → funders who care about health + AI + social impact.